In Many cases the chose of what mail/group product is driven by "Group Think Syndrome". Eg well we had X were i was before we should run that. i do have a Zimbra server were i work, and i have had no issues personally with it. and given a choice if i was to run it personally i would. But in some cases some of our users have found "issues" with the product. A lot of those issues relate to "its not the Microsoft product" it should do X...
just my $0.02 worth ... ----- Original Message ----- From: "Brian Friday" <[email protected]> To: "SoCal LUG Users List" <[email protected]> Sent: Sunday, September 21, 2008 9:06:55 PM GMT -08:00 US/Canada Pacific Subject: Re: [LinuxUsers] Exchange drop-in replacement now opensource At the risk of continuing the flame war. Why don't we just do the following, Open Standard Microsoft Standard: ======================================= IMAP MAPI POP3 MAPI LDAP Active Directory "ldap compatible" ODF OpenXML SPF* CallerID* * Note I think both of these suck but thats just my opinion While I didn't put a lot above I think if you look at the past you will see that at every point where Microsoft could contribute to a group effort they refused to do so when they could get away with it (check out groklaw odf vs openxml). If they had to work in a group they did so very unwillingly and under significant protest/dragging of feet. For them ideas are good but only if they control them. It makes sense really because this is not about them hating ideas only that they want to control the business market... In terms of Zimbra, souped up imap really is a bad moniker. What they have done like many others is create a method which allows end users to use what they want while not requiring that you be locked into Microsoft. If you look into the annuals of history I would be willing to bet that before Exchange was a developer who created this initial idea that either Microsoft bought outright or creatively borrowed.... In general I would have to say that Exchange is the souped up imap, but it is like the Seinfeld soup nazi soup. They control, the when, where and how you can access it including the tools you should use. They do not interoperate, you do and if you want to send email to a non Microsoft person the fact you can do that now is a bug not a designed feature. Ultimately IMHO its about Lock in vs Open Standards, they go with Lock in because they do not understand why open standards are useful and Lock in methods have a cheaper initial cost of ownership with increasingly expensive ongoing costs. Open standards require a larger initial cost of ownership but diminishing or non-existing upkeep costs. *shrug* It is an argument for which we will never come up with a reasonable and equitable solution. - Brian On Sep 21, 2008, at 8:15 PM, David Kaiser wrote: > Okay I see what you are referring to as 'souped-up IMAP' as the Zimbra > type solution, where there is a one-stop one-package install. I > personally have never worked with any of those. I thought you were > implying that a properly designed system such as I've built was > considered as 'souped-up IMAP'. (And I see a properly designed e-mail > system as WAY more than souped-up IMAP, obviously) > > When I've made the soup, it's been pretty tasty however - with real > solid anti-spam and anti-virus, clustered storage on the backend > (over a > completely open protocol to boot), we could actually move users and > even add more storage servers on the backend without any downtime for > the existing users, etc. > > What I would really like to see is if there was a product, such as the > aforementioned on on your first post, that was completely capable of > speaking all the right protocols for Outlook clients, but on the > backend > supported a real LDAP system, supported open-source mail processing, > routing and storage systems such as Postfix and maybe Cyrus IMAP etc. > I still think that would be very viable. Essentially replacing the > 'Exchange' piece of the equation with something that a real admin can > really do awesome stuff with. Eventually, those protocols should be > exposed to Thunderbird, evolution, etc. so that Outlook is only one of > many choices in the field. It would be neat to actually market > something open and flexible - if it contained every combination of > those > little features that Outlook users can't seem to change away from. > > > > Roger E. Rustad, Jr wrote: >> David Kaiser wrote: >> >>> I completely disagree with you. I am one of the biggest critics of >>> Exchange BECAUSE I have been one that worried about directory >>> services >>> for a large organization, >>> >> >> All I said was something I have found in my experience. >> >> I still maintain that most of the biggest critics I have dealt with >> on >> this issue are those who have not really had to seriously deal with >> directory service nuances for large enterprises. You are but one of >> many >> of those critics. >> >> I also still maintain that NT's way of doing things has its place in >> business environments. It is horribly limited compared to other >> solutions, particularly messy campus ones (LSU might be a good >> example >> that Brian can comment on), but it meets the immediate and future >> business needs of lots of companies. >> >> >>> I can say that your analogy shows how little you understand about >>> non-Exchange environments. There is no such thing as a souped-up >>> IMAP >>> server. A souped-up LDAP server, on the other hand, kick's >>> Exchange + >>> Active Directory's @$$ - left, right and center. With the LDAP >>> system >>> that I was setting up at LLU for example, you could do all directory >>> functions from any OS, Linux Windows Mac Solaris whatever. You >>> could >>> authenticate for all kind of services, not just e-mail, manage your >>> password, maintain certain directory information, from ANY OS. Do >>> _that_ with an Active Directory setup! >>> >> >> Ok, if you're wanting to say you know more than me Directory >> Services, >> I'll give you that. There are IMAP servers (their primary role) >> where >> admins have hacked on various other programs to integrate it with >> Exchange. A lot of other products (Zimbra) are very cool, although >> there >> are still lots of things that they can't do. Enterprises that need >> those few (and often times stupid) features that MS offers "cannot" >> change to these (and I still maintain) "souped up IMAP" servers. >> >> Souped up stuff comes with a price. It may run circles around other >> solutions, but I've seen time and time again that the hotshot guy >> who is >> really good at souping up things quite often is the same guy who >> likes >> flipping the CEO the bird and has better things to do than justify to >> others why he choose to make the decisions he did. >> >> >>> I see it as all about vendor lock-in. Do you want to use a product >>> which dictates every other components in the system? >>> >> >> Sometimes, yes. >> >> Sometimes I don't care. And sometimes I really do care, but the >> client >> doesn't care and isn't willing to pay the price. >> >> It all depends on existing infrastructure, existing IT expertise, and >> anticipated future use. Attitudes, skill sets, and budgets all >> sometimes >> compel you to go with an "inferior" vendor lock-in solution. >> >> >>> Do you want to use Active Directory, where you have to run it on >>> Windows, have to authenticate from a Windows host, have to use >>> only a >>> Windows host to change your password? Or do you want to use >>> something >>> based on a standard protocol such as LDAP that does every one of >>> those >>> functions better and allows you to do this from any OS? >>> >> >> I agree that MS AD is not as extensible as RADIUS, and I agree that >> it's >> a vendor lockin. But those don't matter for many companies. >> >> >>> Do you want to use Exchange, which is an e-mail server that only >>> works >>> with Active Directory, only allows for up to 12 e-mail rules per >>> user, >>> limits the type of storage options (proprietary mailbox disk format >>> anyone?) etc... or do you want to have your choice of any of the >>> standards-based e-mail servers that allow for users to fully >>> customize >>> their delivery rule options, and allow the administrator to have >>> full >>> control over how the mail is stored? >>> >> >> Sometimes, yes. >> >> I don't always have power users, and I know that the company's budget >> compels them to hire a lower end administrator. Sometimes when I >> have to >> bid out the job, I'm not quite sure how many hours these other >> solutions >> will take. >> >> I know how long can almost tell you to the minute how long certain >> tasks >> will take on Microsoft. Some of these other open source solutions >> are >> open variables, and I don't know them as well, and because I don't, I >> don't bid them. Sure, I could figure them out, but there's little >> chance >> that the company will want to inherit that solution afterwards. >> >> Also, sometimes I do not want certain admin to have full control over >> certain features. Some admins cannot handle sharp tools and we >> have to >> limit the damage that they can do, while trying to maximize their >> effectiveness by creating strict roles and not letting them stray too >> far from those roles. >> >> >>> Well, eventually at LLU, and also with every story you've told about >>> your Exchange setups, the choice was not about selecting a system >>> which >>> offered flexible options to ensure that the long-term maintenance >>> was >>> manageable - it is always about ease of initial install. The real >>> reason Exchange gets installed over a standards-based e-mail >>> system is >>> that the decision is left to pinheads that think clicking on >>> SETUP.EXE >>> and running a wizard a few times gives them the optimal system. >>> >> >> Yes, that was a huge part of it. >> >> Sure, the vast majority of Windows admins are in my opinion (as you >> put) >> "pinheads". The Linux admins can be a different sort of pinhead. How >> many Linux admins do you know who have tweaked out everything with no >> documentation? >> >> Pick your poison, but it's difficult to claim that either way is >> categorically superior or inferior to the other. >> >> Good NT admins know about official MS ways of doing things, know the >> right registry tweaks, and know what MS is good for (and not good >> for). >> >> >>> Yes, if you really manage directory services for a large >>> organization - >>> you'll see that true LDAP directories and multi-OS support wins >>> out over >>> "ease of install but hard to fix later" Exchange. >>> >> >> I see this in almost any solution. >> >> I work for a company that sells products that are (in one sense of >> the >> word) more difficult than Cisco to configure. (Cisco is, in some >> ways, >> the "Microsoft" of the networking world). >> >> Most of the time, ease-of-use comes with the price of extensibility. >> Pick your poison. >> >> >>> If this Exchange "drop-in replacement" application really serves >>> Calendars to Outlook clients and really does work with open and >>> standards based protocols on the backend, LDAP, IMAP, etc. then >>> to a >>> professional sysadmin, it is certainly a viable replacement for >>> Exchange, if not a preferred one. >>> >> >> In your opinion, which one solution does this? >> >> If not one solution, which home-rolled solutions? >> _______________________________________________ >> LinuxUsers mailing list >> [email protected] >> http://socallinux.org/cgi-bin/mailman/listinfo/linuxusers >> > _______________________________________________ > LinuxUsers mailing list > [email protected] > http://socallinux.org/cgi-bin/mailman/listinfo/linuxusers _______________________________________________ LinuxUsers mailing list [email protected] http://socallinux.org/cgi-bin/mailman/listinfo/linuxusers
